problem with adding constraint to the model

回答済み

ahmad alanaqreh

2021年05月12日 11:12

• 
• 
• 

Hi 

I have a model which contains variables and parameters in this way below

with gp.Model ( "masterproblem" ) as model:

    Z_var = model.addVar ( name="Z")
    y = {}
    for a in range ( 1, n + 1 ):
        y[a] = model.addVar ( lb=0, ub=1, name="y%d" % (a))

    alpha = defaultdict ( lambda: 1, {num_Cuts: 1, } )

    for (i, j) in E:
        if i != s and j != s:
            beta = defaultdict ( lambda: 1, {(num_Cuts, i, j): 1, } )

    for v in range ( 1, n_prim + 1 ):
        for w in range ( 1, n_prim + 1 ):
            delta = defaultdict ( lambda: 1, {(num_Cuts, v, w): 1, } )

    for v in range ( 1, n_prim + 1 ):
        segma = defaultdict ( lambda: 1, {(num_Cuts, v): 1, } )

    for (i, j) in E:
        if i != s and j != s:
            mu = defaultdict ( lambda: 1, {(num_Cuts, i, j): 1, } )

    # Set global sense for ALL objectives
    model.ModelSense = GRB.MINIMIZE
    # Set up primary objective
    model.setObjective ( Z_var )
    model.write ( 'MP.lp' )

now I have a while loop and every time I am adding a constraint to the model as follows :

while "condition":
        MP_model = gp.read ( "MP.lp" )
        k=1
        lhs=Z_var
        rhs= (2 * alpha[k] + quicksum ( y[i] * beta[k, i, j] for (i, j) in E if i != s and j != s )
                          + 2 * quicksum ( y[v] * delta[k, v, v] for v in range ( i, n_prim + 1 ) if v != s )
                          + 2 * quicksum ( y[v] * segma[k, v] for v in range ( 1, n_prim + 1 ) if v != s )
                          + 2 * quicksum ( mu[k, i, j] for (i, j) in E if i != s and j != s ))
        sense=GRB.GREATER_EQUAL
        name=k
        MP_model.addConstr ( lhs, sense, rhs, name )
        MP_model.update ()
        k+=1

but I got an error message that 

 MP_model.addConstr ( lhs, sense, rhs, name )
File "src\gurobipy\model.pxi", line 3376, in gurobipy.Model.addConstr
File "src\gurobipy\model.pxi", line 3206, in gurobipy.Model.__addConstr
gurobipy.GurobiError: Variable not in model

0

• 正式なコメント

  

  Simranjit Kaur

  • Gurobi Staff

  2025年06月23日 02:55 編集済み

  This post is more than three years old. Some information may not be up to date. For current information, please check the Gurobi Documentation or Knowledge Base. If you need more help, please create a new post in the community forum. Or why not try our AI Gurobot?.
• 

  Jaromił Najman

  • Gurobi Staff

  2021年05月12日 14:11 編集済み

  Hi Ahmad,

  Variables \(\texttt{Z_var}\) and \(\texttt{y}\) are part of the \(\texttt{model}\) object and not of the \(\texttt{MP_model}\). In order to access them in the most convenient way, you could add

  MP_model = gp.read ( "MP.lp" )
  y = MP_model.getVars() # get all vars as list
  
  Z_var = 0
  for v in y:
  # this only works because all variable are named y# except for one Z
    if v.VarName == 'Z': 
      Z_var = v # save variable Z to Z_var
      y.remove(v) # remove variable Z from the list of y variables
      break

  To be 100% certain, you could also check whether the \(\texttt{y}\) list is in the correct order.

  Best regards,
  Jaromił

  1
• 

  ahmad alanaqreh

  2021年05月12日 18:36 編集済み

  • 
  • 
  • 

  Dear Jaromil 

  Thank you so much for your help, it's working perfectly :)

  I have another question if you don't mind, I have a model and I am solving the model if it is infeasible, I want to get the unbounded value of each constraint, I do the following :

  SP_model.getConstrByName ( 'one' ).FarkasDual

  and if there is a solution I want the dual value of the constraint :

  SP_model.getConstrByName ( 'one' ).Pi

  I don't know if this is the right way because for Pi it is working normally but with Farkas dual, I got this error message :

  alpha[k] = SP_model.getConstrByName ( 'one' ).FarkasDual
    File "src\gurobipy\constr.pxi", line 67, in gurobipy.Constr.__getattr__
    File "src\gurobipy\constr.pxi", line 95, in gurobipy.Constr.getAttr
    File "src\gurobipy\attrutil.pxi", line 100, in gurobipy.__getattr
  AttributeError: Unable to retrieve attribute 'FarkasDual'

  Thanks a lot 

  0
• 

  Jaromił Najman

  • Gurobi Staff

  2021年05月14日 07:35 編集済み

  Hi Ahmed,

  FarkasDual values are only available if you set the parameter InfUnbdInfo to 1. This is done to avoid the rather expensive computations connected to FarkasDual values. In you particular case, you will have to set

  SP_model.setParam("InfUnbdInfo",1)
  SP_model.optimize()

  Best regards,
  Jaromił

  1

投稿コメントは受け付けていません。